Project Summary/Abstract The utility of nuclear magnetic resonance (NMR) comes largely because it produces a readily interpretable ?molecular fingerprint?; this feature (and the lack of ionizing radiation) also provide important advantages to its clinical cousin, magnetic resonance imaging (MRI) over other whole-body imaging modalities. Unfortunately, sensitivity is almost always a limitation. The conventional solution, going to ever higher field strengths, is limited by extraordinary costs (?$10M) and the complexity of using frequencies with wavelength comparable to body size. More recently, an exciting new frontier involves ?hyperpolarization? methods (mostly dissolution dynamic nuclear polarization or d-DNP) which can increase the observable signal by 104-105 in virtually any organic molecule. Demonstrated applications ranging from monitoring reaction dynamics to clinical imaging, mostly of pyruvate and its enzymatic derivatives in prostate cancer. However, the d-DNP apparatus is complicated, expensive (about $2.5M for the commercial SPINLabTM clinical systems) and slow (it takes many minutes to polarize). A new family of hyperpolarization methods (called ?extended SABRE? or X-SABRE) uses para- hydrogen to drastically alter this picture; polarization is achieved in seconds, on about 100 different molecules to date with very different structures, using an apparatus that can be built for a few thousand dollars. Still, this is hyperpolarization on a laboratory scale, not a clinical one: we typically produce 1-10 mg of hyperpolarized samples, as opposed to roughly 1 g in the SPINLabTM system. The focus of this R21 is to scale these methods up to direct competition with d-DNP on the clinical scale, while keeping cost and speed advantages. Our recent work has shown that X-SABRE is limited not by sample volumes, but by the limited solubility of para-hydrogen and challenges with mixing. We propose to solve this with cosolvent-enhanced supercritical CO2 which has suitable physicochemical properties and is highly inert, non- toxic and inexpensive. Hydrogen is miscible in supercritical CO2, and this combined with low viscosity and high diffusivity should allow vastly more effective use of available hydrogen. This will also permit the use of water- insoluble catalysts to hyperpolarize water soluble targets, immediately improving biocompatibility. The expected outcome of these aims is a very general hyperpolarization strategy which retains enormous advantages in cost, complexity and speed compared to dissolution DNP, and permit innovative approaches such as continuous injection of hyperpolarized agents. In addition, since the theoretical SNR of hyperpolarized MRI is essentially independent of field strength, eliminating the staggering cost of the d-DNP approach would make much less expensive (even portable) low-field clinical MRI instrumentation feasible.